A hybrid sentiment analysis model to detect racist tweets using lexicon-based sentiment analysis and a support vector machine algorithm

Author Identifier (ORCID)

Laizah Sashah Mutasa: https://orcid.org/0000-0003-1377-2862

Abstract

Sentiment analysis (also known as opinion mining) is a natural language processing (NLP) technique for determining data’s positive, negative, or neutral nature. The rise of social media platforms such as X (formally Twitter) and Facebook have become great arenas for discourse on racism and mediums of racism ideologies. This study utilized a hybrid sentiment analysis to detect racist tweets using lexicon-based sentiment analysis and a Support Vector Machine. The models’ success in accurately classifying sentiments related to racism highlights its potential for broader applications in the analysis of other social issues. Furthermore, this study contributes to the ongoing discourse on combating racism in the digital age. By shedding light on the sentiments expressed online, it provides valuable insights that can inform policy decisions, advocacy efforts, and public awareness campaigns. The findings underscore the importance of addressing racism not just in the physical world but also in the digital sphere, where harmful ideologies can spread rapidly and widely.

Document Type

Conference Proceeding

Date of Publication

1-1-2026

Volume

2721 CCIS

Publication Title

Communications in Computer and Information Science

Publisher

Springer

School

School of Business and Law

Comments

Kyei, E. A., Asare, J. W., Modey, P., Ujakpa, M. M., Mutasa, L. S., Freeman, E., Brown-Acquaye, W. L., Forgor, L., & Koi-Akrofi, G. Y. (2026). A hybrid sentiment analysis model to detect racist tweets using Lexicon-Based sentiment analysis and a support vector machine algorithm. In Communications in Computer and Information Science (Vol. 2721, pp. 365–389). Springer. https://doi.org/10.1007/978-3-032-12313-8_29

Copyright

subscription content

First Page

365

Last Page

389

Share

 
COinS
 

Link to publisher version (DOI)

10.1007/978-3-032-12313-8_29